Western Digital working on 20,000 RPM Raptor

[DFI pit bull]

Hope these new raptors have a larger cache, and I also hope the 20k spin really is true.

[Nanometer]

The bumps in speed the raptor line has been receiving is nothing short of pathetic. They only research technology when they think it will help markey shares... pigs.

[Whokidmo]

firstly, 20k raptors plus maybe 32mb cache on a single 150gb platter in 2.5inch drive format will be a PERFECT competitor for the SSD market... they are doin this R&D not to sell HDDs for 2 years and then leave the market... that would be the worst business decision ever... they are doing is at a competitor for the SSDs... a veloci-raptor is nearly comparable to a CHEAP SSD even better in write speeds... now imagine 150gb raptor-PRO for like 250... i'd buy it... hell i'd buy 4 just cause of the upgrade plan WD has so I can upgrade to 1TBs in the future and buy SSDs then... think like a business man not idealism... in 3 years i will have built 2 computers buy then... so will have most of you... i'm sorry i dont see a 128gb or 256gb ssd costing less then 5 bucks per gb in the next 18months especially the nice ones... that will be comparable to raptor... once the average... NOT BUDGET... maybe even performance SSDs come down to about 2 bucks a gbs then i will buy... until then... im buyin these raptor-PROs.... and i'll upgrade them to 1TBs when they become obsolete

BTW i got some of my info about SSD speeds here
http://www.legitreviews.com/article/715/1/

secondly... the fact that this is not possible or will not happen... I agree with the other sides.. 300+ MB/s on sata2... plus im not sure but someone said they saw 20k server HDDS before...

think about for 2 seconds... if u have a REALLY REALLY nice interior car but it can only go maximum 20MPH or u can run 20MPH forever... which would you pick? theres no difference... that the same with SSD and these 20k... THEY WILL NOT ONLY BEAT THEM IN PRICE BUT MANY TIMES BE EQUAL OR BETTER IN PERFORMENCE...

PS if u bought a veloci-raptor dont worry just register on WD's site and you can upgrade when they come out...

[Slay0r]

think about for 2 seconds... if u have a REALLY REALLY nice interior car but it can only go maximum 20MPH or u can run 20MPH forever... which would you pick? theres no difference... that the same with SSD and these 20k... THEY WILL NOT ONLY BEAT THEM IN PRICE BUT MANY TIMES BE EQUAL OR BETTER IN PERFORMENCE...

Just.... no. Nothings stopping you from buying them though.

[RealTelstar]

interesting, but they should have made it 2 years ago, and the current velociraptor 4 years ago.

[Iconyu]

Has anyone pointed out yet that Western Digital don't really have the connections or the back ground to jump on the SSD train. Almost everyone making SSD's can either make their own memory or have connections with memory companies with a lot of volume going through their hands. Just saying to WD that the hard drive is dead, go make solid state is like telling them to move into the Tape Backup market.
Just because the hardware involved does the same task, doesn't mean WD, Seagate and others can just step forward and do the job well. Honestly, look at the companies stepping into this market and there's no comparison at all.

[rcofell]

Yes, it will.

Why do people keep thinking that the SATA II bus won't be sufficient for some time to come (or also that the current SATA I spec isn't good enough for traditional non-raptor/non-ssd drives)?

SATA II runs at 3.0Gb/s, or 375MB/s. Even if this 20krpm drive TRIPPLED current drive speeds, it still couldn't saturate the SATA II bus.

Yeah, I'm quoting you two for the same thing as above. 375MB/s possible on SATA II... even with the fanciful *projections* of SSD manufacturers (and really, we all know we have to reduce the numbers they say they may hit in a year or two, right?) we still won't saturate it until well into 2010.

Just a correction, plus thoughts of my own.

The bandwidth of the SATA II is a little misleading, yes the actual bus runs at 3.0Gb/s, but that's only the raw capacity. The thing is it uses an 8b10b encoding scheme, meaning an 8-bit chunk of data being sent across the bus is first mapped to a 10-bit codeword, then transmitted across the channel and evaluated once it reaches the other side; this is done so that the clock data can be recovered without needing a separate signal line(less lines good). That means no matter what the maximum amount of real data crossing the bus is only 80% of the raw, 300MB/s in this case. It should also be noted that this doesn't include any protocol overhead either.

Now for a bit of personal opinion. Remember, the goal is to upgrade the bus before it becomes the bottleneck, but yes for the near future things should be fine. Although it should be realized it tends to take awhile for these standards to be created and then adopted by the market. For HDDs the low sustainable speeds won't be an issue for quite some time, but if they somehow come up with a larger/faster cache that has a better prefetch algorithm then that's where the limitations would probably show first.

As for SSDs, depending on how things progress it could be a problem, but not until around 2010-ish like you said. This could either be from a new form of non-volatile memory or else a different internal architecture of the drive. At least with SSDs being solid state based there's a much larger opportunity to come up with new and innovative designs, doing something similar to striping where all of the memory chips are accessed in parallel to increase throughput could be done. I'm kind of curious how the different manufacturers implement things now

With the projected move of SSDs into the enterprise the technology should hopefully progress quickly. Also considering when it comes to the enterprise market where SCSI(and now SAS) drives are used in sizes lower than we're accustomed to with SATA, I imagine SSDs would be a good shoe-in. Some of the big players with SAN offerings have begun testing the waters, and I found it interesting that IBM will probably join in too.

[saint-francis]

Aren't the Samsung Spinpoint F1 drives faster than raptors any way? WTF? They are cheaper, larger and faster. Just get a couple of those and RAID them. When they get cheaper in a year or two get a couple more and add them to the array. By the time they die companies will be defecating SSD's.

[Star_Hunter]

http://techreport.com/articles.x/14583/1

Raptor is a little faster in avg read and write much has much lower access time, did you even look at reviews before you said that?

[Helmore]

Now for a bit of personal opinion. Remember, the goal is to upgrade the bus before it becomes the bottleneck, but yes for the near future things should be fine. Although it should be realized it tends to take awhile for these standards to be created and then adopted by the market. For HDDs the low sustainable speeds won't be an issue for quite some time, but if they somehow come up with a larger/faster cache that has a better prefetch algorithm then that's where the limitations would probably show first.

SATA 6.0 Gbit/s is already on the road map, although I've got no idea how far they are with the specs and planned roll out.

Aren't the Samsung Spinpoint F1 drives faster than raptors any way? WTF? They are cheaper, larger and faster. Just get a couple of those and RAID them. When they get cheaper in a year or two get a couple more and add them to the array. By the time they die companies will be defecating SSD's.

Nope, those Spinpoint F1s aren't faster than the new VelociRaptors. They are faster than the former Raptors in read and write speeds, but they still can't match them in access times. The new VelociRaptors are quite an improvement over the original Raptors in both read and write speeds as well as access time and are faster than the Spinpoint F1s on both these 'categories'. Oh and access time is one of the most important things for a HDD to make it really noticeably quicker in normal use, except for file transfers, it will make your PC snappier and boot/load-programs faster.
This last criteria is one of the strongest points of a SSD BTW.

[YukonTrooper]

I have a funny feeling WD is wasting their money on this, unless they can get them out the door quickly.

[Serpentarius]

i assume they'll have to change sata2 as well .. since it's the bottle neck ...
maybe sata3?

[WhiteFireDragon]

wow i hope noise, heat, or vibration won't be a problem

[DFI pit bull]

interesting, but they should have made it 2 years ago, and the current velociraptor 4 years ago.

Yes I totally agree.

My raptors are the only thing that have stayed in my rig the last 6 years despite the several upgrades.

[Helmore]

i assume they'll have to change sata2 as well .. since it's the bottle neck ...
maybe sata3?

SATA 3.0 Gbit/s will start to become a bottleneck when they start making 30k rpm raptors or raptors with a lot higher storage density, like a 1TB 2 platter 2,5 inch 10k rpm VelociRaptor. SATA 3.0 Gbit/s (or SATA 2 as you like to call it, but which does not exist) will be sufficient until drives like that start to apear, or once we start having SSDs with more than 250 MB/s read or write speeds.

[mikeyakame]

Why are we concerned about SATA300 / SAS limits.....SAS2/SATA600 has already been specified, but neither of the forementioned standards are bottlenecks. While we develop hard drives capable of doing burst reads for 1-2Gbit Network cards, we still bottleneck on the PCI bus no matter how fast the drives we throw at a computer.

The bus itself isn't the problem, it's the interconnect with the CPU/Ram, and mix that with multiple devices on the bus saturating it simultaneously, the theoretical limits soon cross our paths much earlier than we ever expected. This isn't a big problem for Joe home user, but in Enterprise environments it's an ongoing bottleneck. In many situations Ram Disks are the only sustainable way to deal with such large data throughput. While PCI-E 2.0 should help a lot with this, it will be quite some time before we see implementations in IO controllers in an Enterprise environment or even home.

It's too late for 20K rpm drives to make much difference. It's just WD trying to grasp at a ledge to slow down and reduce the impact of the eminent plummet. This should have happened years ago, not now. Before we lacked bus speeds, now we lack hard drive speeds. Solid state drives appear to bridge that gap.

Building these drives on the SATA interface is insane if that is their intention (cost effective none the less), SAS is at least extremely scalable. There would need to be a rapid acceptance of SAS to make any kind of 20K raptor a viable decision for a customer who wants to make use of what it offers. This would be terribly expensive, as why SCSI based technologies are still far out of the reach of mainstream.

[Super strokey]

I just wanna see incredibly fast SSD drives that are used for OS and other high use programs and huge (like 10 TB+ HDD's) where things like media and what not are pulled off of to use when ever.

[Kunaak]

anyone remember Multi Laser CD Drives? the 72x CD drive...

technically, It isnt possible to do, cause anything spinning near 50X exponentially increases its chance at shattering to dust.
but with multi lasers reading various points on a disc at the same time, they not only were able to slow the disc down greatly, but also increase its performance up to 72x.

I wonder if the same general idea could some how work in a hard drive.

72x drives died out cause of price, and only 1 company picked up the technology.
but the idea is still solid.

[EnJoY]

anyone remember Multi Laser CD Drives? the 72x CD drive...

technically, It isnt possible to do, cause anything spinning near 50X exponentially increases its chance at shattering to dust.
but with multi lasers reading various points on a disc at the same time, they not only were able to slow the disc down greatly, but also increase its performance up to 72x.

I wonder if the same general idea could some how work in a hard drive.

72x drives died out cause of price, and only 1 company picked up the technology.
but the idea is still solid.

Yamaha made a 72x CD-ROM drive, that thing was beast! I remember reading the Maximum PC article.

[[XC] Synthetickiller]

Yamaha made a 72x CD-ROM drive, that thing was beast! I remember reading the Maximum PC article.

I have one. I need to dig it up. It doesn't work. It failed 6 months after getting it. I keep it for historical reasons. I might see if it works just for kicks, although I don't see it coming back to life anytime soon.

[Soulburner]

anyone remember Multi Laser CD Drives? the 72x CD drive...

technically, It isnt possible to do, cause anything spinning near 50X exponentially increases its chance at shattering to dust.
but with multi lasers reading various points on a disc at the same time, they not only were able to slow the disc down greatly, but also increase its performance up to 72x.

I wonder if the same general idea could some how work in a hard drive.

72x drives died out cause of price, and only 1 company picked up the technology.
but the idea is still solid.

I've wondered this before. Why can't we have multiple heads reading from the same platter? Sure it would take some logic for coordination, but it seems so simple there has to be a reason it wasn't done years ago.

[EnJoY]

I have one. I need to dig it up. It doesn't work. It failed 6 months after getting it. I keep it for historical reasons. I might see if it works just for kicks, although I don't see it coming back to life anytime soon.

Wicked cool. I wonder why they don't implement that technology into current DVD-RW/RAM drives to improve read/write speeds?

[[XC] Synthetickiller]

Wicked cool. I wonder why they don't implement that technology into current DVD-RW/RAM drives to improve read/write speeds?

When my dad was alive, he asked the same question when then 72x cd drive came out. He thought that dvd and every sequential tech to come out would use something that face b/c its logical.

I really need to pull it out. Now I have to go digging through a room of old computer crap to find it and take some pics just b/c. Come to think of it, I hope I didn't throw it out. I'll have to go on a journey through the room crap just to find it.

[[XC] Synthetickiller]

I figured I owed you guys. Here's two pics of the drive. I dare not take more since my only camera is a samsung blackjack II. I hate cell phone cameras, but I need to save for something good like a nikon d40.

Here's the Kenwood True-x 72x drive.

As you can see, its old. 1999!


Front!


EDIT:I plugged the drive into an old amd set up I have and windows won't see the drive. It opens fine. I don't hear it spin up. Makes me sad this baby won't fire up or been seen by winblows.

[afX]

This has got to be a joke. Mechanical drives are old technology. By the time they bring this to market, SSDs will be commonplace.

hate to tell you bud, core 2 cpus are still NOT commonplace, forget about ssds.